Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,37 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# Dataset Information
|
| 6 |
+
|
| 7 |
+
TAL_OCR_Composed_37K is a composed set of TAL OCR datasets. The original images and labels are downloaded from TAL website (https://ai.100tal.com/dataset), includes mainly
|
| 8 |
+
|
| 9 |
+
- K12 Handwritten Chinese texts, English texts, Math formulas.
|
| 10 |
+
- Printed K12 materials.
|
| 11 |
+
|
| 12 |
+
Images are tiled randomly to have a more compact view. There are total 32K images and text pairs after processing:
|
| 13 |
+
|
| 14 |
+
- TAL_OCR_CHN/composed (645 images)
|
| 15 |
+
- TAL_OCR_ENG/composed (399 images)
|
| 16 |
+
- TAL_OCR_MATH/composed (420 images)
|
| 17 |
+
- TAL_K12_PRINT/composed (17109 images)
|
| 18 |
+
- HME100K/composed (18625 images)
|
| 19 |
+
|
| 20 |
+
# Label format
|
| 21 |
+
|
| 22 |
+
```
|
| 23 |
+
{"image": "path_to_image", "ocr" : "text", "data_name": "xx"}
|
| 24 |
+
```
|
| 25 |
+
|
| 26 |
+
# Citation
|
| 27 |
+
|
| 28 |
+
Credits to the TAL team (https://github.com/Phymond/HME100K).
|
| 29 |
+
|
| 30 |
+
```
|
| 31 |
+
@article{yuan2022syntax,
|
| 32 |
+
title={Syntax-Aware Network for Handwritten Mathematical Expression Recognition},
|
| 33 |
+
author={Yuan, Ye and Liu, Xiao and Dikubab, Wondimu and Liu, Hui and Ji, Zhilong and Wu, Zhongqin and Bai, Xiang},
|
| 34 |
+
journal={arXiv preprint arXiv:2203.01601},
|
| 35 |
+
year={2022}
|
| 36 |
+
}
|
| 37 |
+
```
|